Submitted by Geyciane Santos (geyciane_thamires@hotmail.com) on 2015-06-18T15:17:30Z No. of bitstreams: 1 Disserta????o - Joacir Marques de Oliveira J??nior.pdf: 5665291 bytes, checksum: 5db2b29d425ab1c0844713edba8edb09 (MD5) Approved for entry into archive by Divis??o de Documenta????o/BC Biblioteca Central (ddbc@ufam.edu.br) on 2015-06-19T20:58:41Z (GMT) No. of bitstreams: 1 Disserta????o - Joacir Marques de Oliveira J??nior.pdf: 5665291 bytes, checksum: 5db2b29d425ab1c0844713edba8edb09 (MD5) Approved for entry into archive by Divis??o de Documenta????o/BC Biblioteca Central (ddbc@ufam.edu.br) on 2015-06-19T20:59:49Z (GMT) No. of bitstreams: 1 Disserta????o - Joacir Marques de Oliveira J??nior.pdf: 5665291 bytes, checksum: 5db2b29d425ab1c0844713edba8edb09 (MD5) Made available in DSpace on 2015-06-19T20:59:49Z (GMT). No. of bitstreams: 1 Disserta????o - Joacir Marques de Oliveira J??nior.pdf: 5665291 bytes, checksum: 5db2b29d425ab1c0844713edba8edb09 (MD5) Previous issue date: 2014-02-14 Among several steps which are necessary for the commercialization of oil, the analysis of well logs plays an important role to estimate the capacity of a well. Traditionally, this analysis is conducted in a semi-automated process which generates graphs of curves used by human experts to analyze and make the reservoir characterization. One goal of this analysis is to classify lithofacies. Lithofacies are lithological units(rocks) that characterize the environment and compositional aspects of the rocks. In order to characterize an oil reservoir, a set of classes of sedimentary rocks occur. This is which is the major reason for the classification of lithofacies. This master thesis investigates the use of automatic classification techniques applied to the problem of classification of lithofacies. The following five classification methods are investigated: Support Vector Machines, k-Nearest Neighbor, Multilayer Perceptron and Logistic Regression. The database investigated consists of samples from three oil wells of the same reservoir in the Amazon State. In addition, the performance of individual classifiers are compared to the combination of the same five classifiers through majority voting. Finally, we will verify whether or not individual classifiers, or ensemble of classifiers, may train using data obtained from one well and accurately classify data from other wells. In order to get these answers, we have run two series of experiments. First, we trained classifiers and test classifiers individually and combined within the same oil well. The obtained results show that Support Vector Machines achieved the best results in two of the three wells, while Multilayer Perceptron ouperformed the other methods in the third well. In the second series of experiments, we trained classifiers with data from a well and them with data from another well, simulating a situation closer to a real application, since we may use a manually classified database to train a classifier, or ensemble of classifiers, in orde to learn the pattern of the reservoir. Then, data from other wells of the same reservoir may be automatically classified. In this test, the ensemble of classifiers outperformed individual classifiers in 4 of the 6 possible combinations. In the two other combinations, the combination by majority vote was the second best. It is also worth saying that in average, ensemble of classifiers was the best option to classify lithofacies. Our results indicate that combining classifiers in a system of majority voting, shows a better performance and better stability of the results. Dentro das v??rias etapas que s??o necess??rias at?? o petr??leo ser comercializado, a an??lise de perfis el??tricos representa papel de grande import??ncia para se estimar a capacidade produtiva de um po??o. A an??lise hoje ?? semi-automatizada e ocorre da seguinte forma: ge??logos especialistas analisam gr??ficos de curvas gerados por um sistema, para ent??o, realizar a caracteriza????o do reservat??rio com base nas an??lises. Um dos objetivos dessa an??lise ?? a classifica????o de litof??cies. Litof??cies s??o unidades litol??gicas (rochas) que caracterizam o ambiente de forma????o e aspectos composicionais das rochas. Para que se forme um reservat??rio de petr??leo, um conjunto de tipos de rochas precisa estar presente, sendo este um dos principais motivos para a classifica????o de litof??cies. Esta disserta????o de mestrado investiga o uso de t??cnicas de classifica????o autom??tica aplicadas ao problema de classifica????o de litof??cies. Ser??o investigados os seguintes cinco m??todos de classifica????o: Support Vector Machines, k Vizinhos Mais Pr??ximos, Multilayer Perceptron e Regress??o Logistica. A base de dados investigada ?? composta por amostras de perfis de tr??s po??os de uma reserva da Amaz??nia. Ser?? ainda comparado o desempenho de classificadores individuais frente ?? combina????o do mesmos atrav??s do voto majorit??rio. Por fim, iremos verificar se o treinamento de um po??o pode ser aproveitado para outro por meio de classificadores individuais e combinados por voto majorit??rio. Para obter essas respostas, fizemos dois tipos de testes. No primeiro, treinamos e testamos os classificadores individualmente e combinados dentro do mesmo po??o. Os resultados apresentados mostraram que Support Vector Machines foi superior em dois dos tr??s po??os, enquanto Multilayer Perceptron, superou os demais m??todos no terceiro po??o. No segundo tipo de testes, treinamos com dados de um po??o e testamos com dados de outro po??o, simulando uma situa????o mais pr??xima do problema real que seria de calibrar os classificadores de uma reserva com um po??o pioneiro e a partir da?? replicar nos po??os vizinhos. Nestes testes, a combina????o de classificadores se mostrou a melhor solu????o em 4 das 6 combina????es poss??veis. Nas duas demais combina????es, a combina????o por voto majorit??rio alcan??ou o segundo melhor resultado. Vale dizer ainda que na m??dia simples o sistema de vota????o majorit??rio, foi a melhor op????o para classificar as litof??cies. Nossos resultados indicam que combinar classificadores em um sistema de vota????o majorit??rio apresenta desempenho superior ao uso de classificadores individuais, al??m de apresentar maior estabilidade.